from sklearn_benchmarks.report import Reporting, ReportingHpo, print_time_report, print_env_info
import pandas as pd
pd.set_option('display.max_colwidth', None)
pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', None)
%matplotlib notebook

print_time_report()

Time report

daal4py_KMeans_short: 0h 0m 1s

daal4py_Ridge: 0h 0m 2s

KMeans_short: 0h 0m 3s

daal4py_LogisticRegression: 0h 0m 5s

daal4py_KMeans_tall: 0h 0m 10s

Ridge: 0h 0m 11s

KMeans_tall: 0h 0m 25s

LogisticRegression: 0h 0m 26s

daal4py_KNeighborsClassifier_kd_tree: 0h 0m 30s

KNeighborsClassifier_kd_tree: 0h 2m 26s

lightgbm: 0h 5m 10s

HistGradientBoostingClassifier: 0h 5m 11s

catboost_symmetric: 0h 5m 22s

xgboost: 0h 5m 29s

catboost_lossguide: 0h 5m 40s

daal4py_KNeighborsClassifier: 0h 6m 15s

KNeighborsClassifier: 0h 34m 16s

total: 1h 11m 51s

print_env_info()

Benchmark environment

{
  "system_info": {
    "python": "3.8.10 | packaged by conda-forge | (default, May 11 2021, 07:01:05)  [GCC 9.3.0]",
    "executable": "/usr/share/miniconda/envs/sklbench/bin/python",
    "machine": "Linux-5.8.0-1033-azure-x86_64-with-glibc2.10"
  },
  "dependencies_info": {
    "pip": "21.1.2",
    "setuptools": "49.6.0.post20210108",
    "sklearn": "1.0.dev0",
    "numpy": "1.21.0",
    "scipy": "1.6.3",
    "Cython": null,
    "pandas": "1.2.5",
    "matplotlib": "3.4.2",
    "joblib": "1.0.1",
    "threadpoolctl": "2.1.0"
  },
  "threadpool_info": [
    {
      "filepath": "/usr/share/miniconda/envs/sklbench/lib/libopenblasp-r0.3.15.so",
      "prefix": "libopenblas",
      "user_api": "blas",
      "internal_api": "openblas",
      "version": "0.3.15",
      "num_threads": 2,
      "threading_layer": "pthreads"
    },
    {
      "filepath": "/usr/share/miniconda/envs/sklbench/lib/python3.8/site-packages/scikit_learn.libs/libgomp-f7e03b3e.so.1.0.0",
      "prefix": "libgomp",
      "user_api": "openmp",
      "internal_api": "openmp",
      "version": null,
      "num_threads": 2
    },
    {
      "filepath": "/usr/share/miniconda/envs/sklbench/lib/libgomp.so.1.0.0",
      "prefix": "libgomp",
      "user_api": "openmp",
      "internal_api": "openmp",
      "version": null,
      "num_threads": 2
    }
  ],
  "cpu_count": 2
}

reporting = Reporting(config="config.yml")
reporting.run()

KNeighborsClassifier: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: algorithm=brute.

	estimator	function	n_samples_train	n_samples	n_features	n_iter	mean_sklearn	stdev_sklearn	throughput	latency	n_jobs	n_neighbors	accuracy_score_sklearn	accuracy_score_daal4py	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.163	0.000	4.896	0.000	-1	1	NaN	NaN	0.530	0.000	0.308	0.000	See	See
1	KNeighborsClassifier	predict	1000000	1000	100	NaN	31.418	0.000	0.000	0.031	-1	1	0.721	0.833	4.398	0.025	7.143	0.040	See	See
2	KNeighborsClassifier	predict	1000000	1	100	NaN	0.206	0.018	0.000	0.206	-1	1	0.000	1.000	0.106	0.002	1.952	0.172	See	See
3	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.145	0.000	5.510	0.000	1	5	NaN	NaN	0.538	0.000	0.270	0.000	See	See
4	KNeighborsClassifier	predict	1000000	1000	100	NaN	27.315	0.542	0.000	0.027	1	5	0.797	0.928	4.510	0.040	6.057	0.132	See	See
5	KNeighborsClassifier	predict	1000000	1	100	NaN	0.221	0.005	0.000	0.221	1	5	0.000	1.000	0.114	0.004	1.941	0.079	See	See
6	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.132	0.000	6.067	0.000	-1	5	NaN	NaN	0.533	0.000	0.248	0.000	See	See
7	KNeighborsClassifier	predict	1000000	1000	100	NaN	41.552	0.000	0.000	0.042	-1	5	0.797	0.702	4.413	0.025	9.416	0.053	See	See
8	KNeighborsClassifier	predict	1000000	1	100	NaN	0.224	0.022	0.000	0.224	-1	5	0.000	1.000	0.101	0.002	2.229	0.224	See	See
9	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.156	0.000	5.138	0.000	1	1	NaN	NaN	0.538	0.000	0.289	0.000	See	See
10	KNeighborsClassifier	predict	1000000	1000	100	NaN	17.276	0.361	0.000	0.017	1	1	0.721	0.833	4.410	0.010	3.918	0.082	See	See
11	KNeighborsClassifier	predict	1000000	1	100	NaN	0.214	0.004	0.000	0.214	1	1	0.000	1.000	0.109	0.002	1.967	0.053	See	See
12	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.145	0.000	5.514	0.000	1	100	NaN	NaN	0.567	0.000	0.256	0.000	See	See
13	KNeighborsClassifier	predict	1000000	1000	100	NaN	27.220	0.443	0.000	0.027	1	100	0.923	0.702	4.399	0.046	6.188	0.119	See	See
14	KNeighborsClassifier	predict	1000000	1	100	NaN	0.229	0.004	0.000	0.229	1	100	1.000	1.000	0.104	0.002	2.202	0.053	See	See
15	KNeighborsClassifier	fit	1000000	1000000	100	NaN	0.143	0.000	5.596	0.000	-1	100	NaN	NaN	0.546	0.000	0.262	0.000	See	See
16	KNeighborsClassifier	predict	1000000	1000	100	NaN	41.481	0.000	0.000	0.041	-1	100	0.923	0.928	4.515	0.026	9.187	0.054	See	See
17	KNeighborsClassifier	predict	1000000	1	100	NaN	0.215	0.023	0.000	0.215	-1	100	1.000	1.000	0.106	0.003	2.018	0.221	See	See
18	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.056	0.000	0.285	0.000	-1	1	NaN	NaN	0.100	0.000	0.562	0.000	See	See
19	KNeighborsClassifier	predict	1000000	1000	2	NaN	24.898	0.876	0.000	0.025	-1	1	0.978	0.982	0.993	0.012	25.062	0.935	See	See
20	KNeighborsClassifier	predict	1000000	1	2	NaN	0.022	0.002	0.000	0.022	-1	1	1.000	1.000	0.005	0.000	4.697	0.625	See	See
21	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.054	0.000	0.296	0.000	1	5	NaN	NaN	0.098	0.000	0.551	0.000	See	See
22	KNeighborsClassifier	predict	1000000	1000	2	NaN	20.869	0.313	0.000	0.021	1	5	0.990	0.986	1.073	0.021	19.457	0.477	See	See
23	KNeighborsClassifier	predict	1000000	1	2	NaN	0.024	0.001	0.000	0.024	1	5	1.000	1.000	0.005	0.000	5.122	0.422	See	See
24	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.052	0.000	0.305	0.000	-1	5	NaN	NaN	0.105	0.000	0.498	0.000	See	See
25	KNeighborsClassifier	predict	1000000	1000	2	NaN	35.147	0.000	0.000	0.035	-1	5	0.990	0.974	0.981	0.014	35.828	0.510	See	See
26	KNeighborsClassifier	predict	1000000	1	2	NaN	0.037	0.003	0.000	0.037	-1	5	1.000	1.000	0.004	0.000	8.321	0.836	See	See
27	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.111	0.000	0.144	0.000	1	1	NaN	NaN	0.096	0.000	1.158	0.000	See	See
28	KNeighborsClassifier	predict	1000000	1000	2	NaN	11.538	0.143	0.000	0.012	1	1	0.978	0.982	0.980	0.017	11.777	0.246	See	See
29	KNeighborsClassifier	predict	1000000	1	2	NaN	0.015	0.001	0.000	0.015	1	1	1.000	1.000	0.005	0.000	3.183	0.265	See	See
30	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.053	0.000	0.301	0.000	1	100	NaN	NaN	0.101	0.000	0.528	0.000	See	See
31	KNeighborsClassifier	predict	1000000	1000	2	NaN	20.848	0.372	0.000	0.021	1	100	0.989	0.974	0.986	0.020	21.134	0.566	See	See
32	KNeighborsClassifier	predict	1000000	1	2	NaN	0.028	0.001	0.000	0.028	1	100	1.000	1.000	0.004	0.000	6.372	0.456	See	See
33	KNeighborsClassifier	fit	1000000	1000000	2	NaN	0.054	0.000	0.298	0.000	-1	100	NaN	NaN	0.095	0.000	0.566	0.000	See	See
34	KNeighborsClassifier	predict	1000000	1000	2	NaN	33.597	0.000	0.000	0.034	-1	100	0.989	0.986	1.070	0.023	31.413	0.662	See	See
35	KNeighborsClassifier	predict	1000000	1	2	NaN	0.035	0.003	0.000	0.035	-1	100	1.000	1.000	0.005	0.000	7.544	0.657	See	See

KNeighborsClassifier_kd_tree: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: algorithm=kd_tree.

	estimator	function	n_samples_train	n_samples	n_features	n_iter	mean_sklearn	stdev_sklearn	throughput	latency	n_jobs	n_neighbors	accuracy_score_sklearn	accuracy_score_daal4py	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.522	0.000	0.032	0.000	1	1	NaN	NaN	0.767	0.000	3.288	0.000	See	See
1	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	0.630	0.015	0.000	0.001	1	1	0.960	0.982	0.623	0.014	1.012	0.033	See	See
2	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.001	0.000	0.000	0.001	1	1	1.000	1.000	0.001	0.000	1.250	0.605	See	See
3	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.573	0.000	0.031	0.000	-1	5	NaN	NaN	0.765	0.000	3.364	0.000	See	See
4	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	0.780	0.021	0.000	0.001	-1	5	0.972	0.982	0.200	0.005	3.903	0.141	See	See
5	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.004	0.000	0.000	0.004	-1	5	1.000	1.000	0.000	0.000	8.281	3.967	See	See
6	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.528	0.000	0.032	0.000	-1	100	NaN	NaN	0.763	0.000	3.312	0.000	See	See
7	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	2.536	0.072	0.000	0.003	-1	100	0.972	0.982	0.200	0.007	12.688	0.562	See	See
8	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.005	0.001	0.000	0.005	-1	100	1.000	1.000	0.000	0.000	12.327	6.124	See	See
9	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.509	0.000	0.032	0.000	1	100	NaN	NaN	0.729	0.000	3.440	0.000	See	See
10	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	4.448	0.106	0.000	0.004	1	100	0.972	0.967	0.109	0.004	40.690	1.703	See	See
11	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.003	0.001	0.000	0.003	1	100	1.000	1.000	0.000	0.000	9.652	5.546	See	See
12	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.661	0.000	0.030	0.000	-1	1	NaN	NaN	0.713	0.000	3.730	0.000	See	See
13	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	0.415	0.010	0.000	0.000	-1	1	0.960	0.982	0.600	0.015	0.692	0.024	See	See
14	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.003	0.001	0.000	0.003	-1	1	1.000	1.000	0.001	0.000	3.183	1.198	See	See
15	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	NaN	2.726	0.000	0.029	0.000	1	5	NaN	NaN	0.722	0.000	3.775	0.000	See	See
16	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	NaN	1.311	0.025	0.000	0.001	1	5	0.972	0.967	0.107	0.003	12.252	0.441	See	See
17	KNeighborsClassifier_kd_tree	predict	1000000	1	10	NaN	0.002	0.000	0.000	0.002	1	5	1.000	1.000	0.000	0.000	5.489	2.790	See	See
18	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.831	0.000	0.019	0.000	1	1	NaN	NaN	0.502	0.000	1.656	0.000	See	See
19	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.030	0.002	0.001	0.000	1	1	0.978	0.988	0.008	0.001	3.941	0.360	See	See
20	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.001	0.000	0.000	0.001	1	1	0.000	1.000	0.000	0.000	4.378	1.768	See	See
21	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.778	0.000	0.021	0.000	-1	5	NaN	NaN	0.500	0.000	1.556	0.000	See	See
22	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.032	0.002	0.000	0.000	-1	5	0.986	0.988	0.001	0.000	25.018	5.096	See	See
23	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.003	0.000	0.000	0.003	-1	5	0.000	1.000	0.000	0.000	15.991	9.148	See	See
24	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.775	0.000	0.021	0.000	-1	100	NaN	NaN	0.493	0.000	1.571	0.000	See	See
25	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.050	0.006	0.000	0.000	-1	100	0.985	0.988	0.001	0.000	39.292	10.662	See	See
26	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.003	0.001	0.000	0.003	-1	100	0.000	1.000	0.000	0.000	14.689	7.970	See	See
27	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.773	0.000	0.021	0.000	1	100	NaN	NaN	0.513	0.000	1.507	0.000	See	See
28	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.052	0.002	0.000	0.000	1	100	0.985	0.985	0.001	0.001	49.917	25.726	See	See
29	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.001	0.000	0.000	0.001	1	100	0.000	1.000	0.000	0.000	4.481	3.184	See	See
30	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.781	0.000	0.020	0.000	-1	1	NaN	NaN	0.525	0.000	1.489	0.000	See	See
31	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.031	0.002	0.001	0.000	-1	1	0.978	0.988	0.008	0.001	4.027	0.389	See	See
32	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.003	0.001	0.000	0.003	-1	1	0.000	1.000	0.000	0.000	16.131	9.023	See	See
33	KNeighborsClassifier_kd_tree	fit	1000000	1000000	2	NaN	0.781	0.000	0.020	0.000	1	5	NaN	NaN	0.541	0.000	1.443	0.000	See	See
34	KNeighborsClassifier_kd_tree	predict	1000000	1000	2	NaN	0.030	0.001	0.001	0.000	1	5	0.986	0.985	0.001	0.000	34.683	12.073	See	See
35	KNeighborsClassifier_kd_tree	predict	1000000	1	2	NaN	0.001	0.000	0.000	0.001	1	5	0.000	1.000	0.000	0.000	4.918	3.123	See	See

KMeans_tall: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: algorithm=full, n_clusters=3, max_iter=30, n_init=1, tol=1e-16.

	estimator	function	n_samples_train	n_samples	n_features	n_iter_sklearn	mean_sklearn	stdev_sklearn	throughput	latency	init	adjusted_rand_score_sklearn	n_iter_daal4py	adjusted_rand_score_daal4py	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	KMeans_tall	fit	1000000	1000000	2	30	0.662	0.0	0.725	0.000	random	NaN	30	NaN	0.366	0.0	1.808	0.000	See	See
1	KMeans_tall	predict	1000000	1000	2	30	0.002	0.0	0.278	0.000	random	0.001	30	0.001	0.000	0.0	8.011	3.422	See	See
2	KMeans_tall	predict	1000000	1	2	30	0.002	0.0	0.000	0.002	random	1.000	30	1.000	0.000	0.0	9.239	3.758	See	See
3	KMeans_tall	fit	1000000	1000000	2	30	0.691	0.0	0.695	0.000	k-means++	NaN	30	NaN	0.308	0.0	2.244	0.000	See	See
4	KMeans_tall	predict	1000000	1000	2	30	0.002	0.0	0.275	0.000	k-means++	0.001	30	0.001	0.000	0.0	7.071	3.481	See	See
5	KMeans_tall	predict	1000000	1	2	30	0.002	0.0	0.000	0.002	k-means++	1.000	30	1.000	0.000	0.0	6.420	7.661	See	See
6	KMeans_tall	fit	1000000	1000000	100	30	7.824	0.0	3.067	0.000	random	NaN	30	NaN	4.060	0.0	1.927	0.000	See	See
7	KMeans_tall	predict	1000000	1000	100	30	0.002	0.0	12.331	0.000	random	0.002	30	0.002	0.000	0.0	6.005	2.982	See	See
8	KMeans_tall	predict	1000000	1	100	30	0.002	0.0	0.014	0.002	random	1.000	30	1.000	0.000	0.0	9.148	5.342	See	See
9	KMeans_tall	fit	1000000	1000000	100	30	7.865	0.0	3.052	0.000	k-means++	NaN	30	NaN	3.810	0.0	2.064	0.000	See	See
10	KMeans_tall	predict	1000000	1000	100	30	0.002	0.0	12.146	0.000	k-means++	0.001	30	0.002	0.000	0.0	5.794	2.583	See	See
11	KMeans_tall	predict	1000000	1	100	30	0.002	0.0	0.014	0.002	k-means++	1.000	30	1.000	0.000	0.0	8.378	3.487	See	See

KMeans_short: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: algorithm=full, n_clusters=300, max_iter=20, n_init=1, tol=1e-16.

	estimator	function	n_samples_train	n_samples	n_features	n_iter_sklearn	mean_sklearn	stdev_sklearn	throughput	latency	init	adjusted_rand_score_sklearn	n_iter_daal4py	adjusted_rand_score_daal4py	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	KMeans_short	fit	10000	10000	2	20	0.321	0.0	0.010	0.000	k-means++	NaN	20	NaN	0.057	0.0	5.675	0.000	See	See
1	KMeans_short	predict	10000	1000	2	20	0.002	0.0	0.144	0.000	k-means++	0.000	20	-0.001	0.001	0.0	2.234	0.841	See	See
2	KMeans_short	predict	10000	1	2	20	0.002	0.0	0.000	0.002	k-means++	1.000	20	1.000	0.000	0.0	7.605	3.631	See	See
3	KMeans_short	fit	10000	10000	2	20	0.104	0.0	0.031	0.000	random	NaN	20	NaN	0.146	0.0	0.714	0.000	See	See
4	KMeans_short	predict	10000	1000	2	20	0.002	0.0	0.142	0.000	random	-0.000	20	-0.003	0.001	0.0	2.455	0.986	See	See
5	KMeans_short	predict	10000	1	2	20	0.002	0.0	0.000	0.002	random	1.000	20	1.000	0.000	0.0	8.508	4.250	See	See
6	KMeans_short	fit	10000	10000	100	20	1.176	0.0	0.136	0.000	k-means++	NaN	20	NaN	0.276	0.0	4.265	0.000	See	See
7	KMeans_short	predict	10000	1000	100	20	0.003	0.0	4.767	0.000	k-means++	0.281	20	0.252	0.002	0.0	1.987	0.329	See	See
8	KMeans_short	predict	10000	1	100	20	0.002	0.0	0.008	0.002	k-means++	1.000	20	1.000	0.000	0.0	8.099	3.576	See	See
9	KMeans_short	fit	10000	10000	100	20	0.342	0.0	0.468	0.000	random	NaN	20	NaN	0.635	0.0	0.539	0.000	See	See
10	KMeans_short	predict	10000	1000	100	20	0.003	0.0	4.713	0.000	random	0.319	20	0.333	0.002	0.0	2.113	0.287	See	See
11	KMeans_short	predict	10000	1	100	20	0.002	0.0	0.009	0.002	random	1.000	20	1.000	0.000	0.0	7.809	2.887	See	See

LogisticRegression: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: penalty=l2, dual=False, tol=0.0001, C=1.0, fit_intercept=True, intercept_scaling=1, class_weight=nan, random_state=nan, solver=lbfgs, max_iter=100, multi_class=auto, verbose=0, warm_start=False, n_jobs=nan, l1_ratio=nan.

	estimator	function	n_samples_train	n_samples	n_features	n_iter	mean_sklearn	stdev_sklearn	throughput	latency	class_weight	l1_ratio	n_jobs	random_state	accuracy_score	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	LogisticRegression	fit	1000000	1000000	100	[20]	16.688	0.0	[-0.07070071]	0.000	NaN	NaN	NaN	NaN	NaN	2.974	0.0	5.611	0.000	See	See
1	LogisticRegression	predict	1000000	1000	100	[20]	0.000	0.0	[41.91637475]	0.000	NaN	NaN	NaN	NaN	0.534	0.000	0.0	0.902	0.428	See	See
2	LogisticRegression	predict	1000000	1	100	[20]	0.000	0.0	[0.15065814]	0.000	NaN	NaN	NaN	NaN	0.000	0.000	0.0	0.386	0.270	See	See
3	LogisticRegression	fit	1000	1000	10000	[27]	1.366	0.0	[-1.5634742]	0.001	NaN	NaN	NaN	NaN	NaN	1.113	0.0	1.227	0.000	See	See
4	LogisticRegression	predict	1000	100	10000	[27]	0.003	0.0	[85.83192654]	0.000	NaN	NaN	NaN	NaN	0.280	0.004	0.0	0.631	0.088	See	See
5	LogisticRegression	predict	1000	1	10000	[27]	0.000	0.0	[15.43839686]	0.000	NaN	NaN	NaN	NaN	0.000	0.001	0.0	0.147	0.081	See	See

Ridge: scikit-learn (1.0.dev0) vs. daal4py (2021.2.3)

All estimators share the following hyperparameters: alpha=1.0, fit_intercept=True, normalize=deprecated, copy_X=True, max_iter=nan, tol=0.001, solver=auto, random_state=nan.

	estimator	function	n_samples_train	n_samples	n_features	n_iter	mean_sklearn	stdev_sklearn	throughput	latency	max_iter	random_state	r2_score	mean_daal4py	stdev_daal4py	speedup	stdev_speedup	sklearn_profiling	daal4py_profiling
0	Ridge	fit	1000	1000	10000	NaN	0.320	0.000	0.250	0.0	NaN	NaN	NaN	0.323	0.000	0.991	0.000	See	See
1	Ridge	predict	1000	1000	10000	NaN	0.012	0.001	6.697	0.0	NaN	NaN	0.133	0.021	0.001	0.574	0.041	See	See
2	Ridge	predict	1000	1	10000	NaN	0.000	0.000	0.752	0.0	NaN	NaN	NaN	0.000	0.000	0.587	0.452	See	See
3	Ridge	fit	1000000	1000000	100	NaN	1.697	0.000	0.471	0.0	NaN	NaN	NaN	0.408	0.000	4.162	0.000	See	See
4	Ridge	predict	1000000	1000	100	NaN	0.000	0.000	4.120	0.0	NaN	NaN	1.000	0.000	0.000	0.708	0.447	See	See
5	Ridge	predict	1000000	1	100	NaN	0.000	0.000	0.009	0.0	NaN	NaN	NaN	0.000	0.000	0.537	0.454	See	See

Gradient boosting: randomized HPO benchmarks

reporting_hpo = ReportingHpo(config="config.yml")
reporting_hpo.run()

Raw fit times

Raw predict times

Smoothed HPO Curves

The shaded areas represent boostrapped quartiles.

Speedup barplots

Speedup curves

import numpy as np
import pandas as pd

df = pd.DataFrame({
    "method": ["fit", "predict", "fit", "predict"],
    "hash": ["0", "0", "1", "1"],
    "score": [None, "0.705", None, "0.802"]
})

df

	method	hash	score
0	fit	0	None
1	predict	0	0.705
2	fit	1	None
3	predict	1	0.802

df.query("method == 'fit'").merge(df.query("method == 'predict'"), on=["hash"], how="inner")

	method_x	hash	score_x	method_y	score_y
0	fit	0	None	predict	0.705
1	fit	1	None	predict	0.802